Noam Chomsky, linguistics genius? Not so fast

Is Noam Chomsky perhaps above my intellectual pay grade? Yes.

Genius? No.

 

Including and above all, not a genius on linguistics.

Neuroscience has shown the brain is NOT "massively modular," therefore undercutting theories of language of Chomsky, Steve Pinker and others. I first noted this point 15 years ago. I also noted, per Wiki's take on him, that he was too much of a rationalist on this issue in general. 

Indeed, Wiki's piece on generative grammar notes that no evidence for deep structures has been found, and that Chomsky has shrugged that off by a fallacious appeal to Galileo (who DID have evidence for his claims), showing that not only is Chomsky too rationalist, but also that, in additional to being anti-behavioralist, he's anti-empiricist.

Related? Back in 2010, per Doug Hofstadter and Gödel, Escher, Bach, I said the mind, including on language, was massively recursive, not massively modular. And, more and more new research, on linguistics in specific and human mental activity in general, is demonstrating this.

And, some researchers argue that Chomsky's version of a universal grammar violated Ockham's Razor even if it actually does explain things.

Next, per this review of Terence Deacon's "The Symbolic Species," even if a universal grammar does exist to some degree, it might not be the major factor in the origin of language.

Related to that, and also undercutting Pinker to a fair degree, neuroscience HAS shown the brain didn't have a massive explosion 60,000 or whatever years ago. In short, language ability evolved gradually and adaptationally, or even exadaptationally. And "one-tenth of a language" is of as much evolutionary value as a planarium's "one-tenth of an eye."

Chomsky also seems to miss the power and role of cultural evolution in this.

 

Even more, of course, he ignores the power and role of plain old neo-Darwinian evolution. It's clear that other upper-level sentient animals (other primates and cetaceans coming first to mind) communicate. Per teaching sign language to some primates, it's clear that they have at least rudimentary skills at symbol manipulation and recombination. And,  of course, from that, clear that language of some sort is not peculiarly or solely human.

 

Those facts all go directly against the claims of Chomsky, a straight humanities guy, that language IS specifically human. From there, to use Dan Dennett's "skyhooks" idea, he seems to have "pegged" without evidence the massively modular brain, etc.

 

Finally, the end of this Wiki piece on transformative grammar he got some massive signal-boosting on early claims to have invented a new world in linguistic study. Related? He got lucky to be attacking behavioralism with his linguistic ideas just as behavioralism was starting to collapse in general. He admits to getting lucky in being the "somebody" MIT needed to fill the faculty position he was hired for. Remember, Noam is purely a humanities guy.

And, even if Chomsky WERE right? What then? He hadn't shown HOW things like deep syntactical structure evolved, and we still don't know that today, whether the brain is massively modular or not.

To be honest, really, on the origins and development of our original use of language, we now little more today than we did 50 years ago.

And, this may remain permanently unknowable.

And, there ARE alternative, newer, scientifically informed takes on the development of language.

 

(Update: And, among them, more and more research shows not only that brains are not massively modular, but that the whole old functional diagram of brains, including the alleged primary function not only of the cerebral cortex's different surface areas, but also separate portions like the cerebellum and amydala, is so out of whack it's probably at Paul's Not.EvenWrong. stage. As part of this, just as we know that "one gene ≠ one phenotypic expression" in both that some expressions need multiple genes coding for them, but more to the case, one gene can be part of coding for several expressions in combo with other genes, so, those functional areas of the brain can express multiple mental workings. This Quanta piece has plenty more.) 

The Truth about Language: What It Is and Where It Came From by Michael C. Corballis
My rating: 5 of 5 stars

Excellent book from the preface on.

Many people know the name of Noam Chomsky, but they may not know that, while he dethroned B.F. Skinner’s behaviorist approach to linguistics, his own theory, which broadly falls into humanist linguistics, has itself become largely passé.

Two major newer schools, with a fair amount of overlap but with distinct emphases, are in the lead today: functionalist and Darwinist schools of linguistics. Michael Corballis comes from the later, though he’s conversant with the former. In the same broad train of thought as a Michael Tomasello, he talks in this book about the likely route for development of human language.

Corballis says straight up that he knew he would butt heads with Chomsky, Gould and others. He rejects Chomsky’s massive modularity of the brain (as does most modern neuroscience) and rejects Gould for saltationist ideas about the origin of language.

Corballis says that he sees normal, incremental neo-Darwinian evolution at work.

Early in part 1, chapter 1, he calls out Chomsky for ignoring most of the vast variation between languages in his attempt to posit a universal grammar. He even QUOTES Chomsky to that effect.

“I have not hesitated to propose a general principle of linguistic structure on the basis of observation of a single language.”

This is basically like the old “spontaneous emergence” idea of maggots in rotten meat, Galen’s claiming the human liver has seven lobes because monkey livers do, or similar.

Now, after refuting Chomsky, what ideas does Corballis offer up?

First is that language probably in part evolved from gestural issues. He notes that human babies point to things just to note them as an object of attention, vs chimps who point because they want.

Next, he notes humans’ ability to mentally time travel. Tis true, he notes, that corvids may not immediately revisit seed caches if they think another of their species has been spying on them, but that’s about it as far as looking to the future among animals. Elephants and primates seem to retain some memory of deceased loved ones, but of itself, that doesn’t reflect mental time travel backward, really. Only humans seem to have that in great degree. This, in turn is part of larger “displacement” in language, moving ourselves spatially as well as temporally. Related to that is that, in English at least, many prepositions can have both spatial and temporal functions.

Beyond that, he postulates that humans (and possibly earlier members of the genus Homo) having third-order theory of mind, vs primates (and presumably, cetaceans) having only second-order TOM, and a restricted and species-specific one at that, is probably a big factor in language development. Language recursiveness and nesting would seem to underscore this.

In all of this, though, Corballis notes that primates have some gesture usage, and that even dogs can recognize specific human words.

Next, it’s off to grammar. After a basic look at parts of speech, Corballis notes how and why, in English and other language, some things like “helping verbs” evolved … and then, in some successor languages, devolved again. As part of this, and the idea that languages in general started as noun-verb only items similar to modern pidgins, Corballis notes the role of cultural evolution.

Corballis ends with his “Crossing the Rubicon” of how he things language began. This starts by summarizing some of his differences with Chomsky on things like internal vs. external language and their function in language development, language as a means of expressing thought rather than thought itself and more.

With that, he notes that to the degree there was a great leap forward, speech, not language, was it. Abstraction was not inherent to speech. Related to that, he says it’s an open question as to whether all current languages evolved from one Ur-language, or if instead, they started evolving after modern Homo sapiens started splitting.

Corballis does admit that, without more evidence, he too is telling a “just-so story,” and it’s nice for him to end on a note of epistemic humility.

Side note: many of his “peregrinations” during the book are interesting, but I think he spends too much time, with repeated returns, to the Aquatic Ape Hypothesis, when it’s but marginally connected to his main theme.

 

Beyond that? Carl Zimmer reports that new research indicates language evolved primarily for communication, and NOT for thinking. Fun sidebar? This is another overturning of Chomsky's claims about language. (I can't say "research," since Chomsky did basically none.) Also, this would tie in with people like Corballis stressing cultural evolution's role in the development of language.

View all my reviews

Chomsky is also wrong on some philosophical issues, such as claiming humans have universal standards of moral justice. On many issues, no we don't. Even what constitutes "murder" is not 100 percent universal. Per my cultural evolution link above, Chomsky's also surely missed its role and power in development of morals. And, speaking of morals, his actions in the Faurisson affair still raise questions.

Sidebar: Deacon's second book, Incomplete Nature, is way controversial. Here's an interesting review. It defends his invention of a number of new words and generally says "good try, but still not actually explained." Deacon was also charged with plagiarism ... charges that might just be true.

He IS right, very much, on media criticism. I'll give him that.

 

Also, Noam's attacks on AI, with claims like it's behavioralist below the elephants and turtles, are laughable, unscientific and anachronistic. They're unscientific because Chomsky admits that his theories of linguistics aren't scientifically founded. They're anachronistic because behavioralism has even fewer devotees than Freudianism. 

I hope that within 20 years Chomskyism on linguistics has as small a crowd of devotees.